A useful AI-Use Disclosure should not say everything, and it should not say nothing. It should say enough to help the audience understand what AI did, whether that use affects trust or reliance, and who remains responsible for the final product.
AI-use disclosure should not turn on whether AI touched the work in some way. That question is too broad to be useful. The better question is what AI did. Did it help brainstorm? Edit? Draft? Analyze? Create media? Influence a decision? Those uses sit on a spectrum, and they do not all require the same kind of disclosure.
AI disclosure sounds simple until you ask what “AI use” actually means. Is brainstorming the same as drafting? Is editing the same as analysis? This post starts a series on why the goal should not be maximum disclosure, but meaningful disclosure that helps people understand what AI did, who reviewed it, and why it matters.
